Transient ischaemic attack, vascular risk factors and cognitive impairment: a case–controlled study

SIR—Cognitive impairment, especially difficulties with temporal orientation and verbal recall, is associated with the increasing number and severity of vascular risk factors (VRFs) such as hypertension and diabetes [1–3] which can result in an associated impairment of the cerebral microcirculation causing white matter volume changes linked to large artery stiffness [4, 5]. These cognitive deficits can be detected by using simple standard screening tools [6] such as the Mini Mental State Examination [7], Montreal Cognitive Assessment (MoCA) [8] and the DemTec [9], and have been shown to be related to the development of both subclinical (mild) or established vascular disorders [7–12]. However, our understanding of the relation between transient ischaemic attacks (TIAs) and cognitive status is incomplete. We hypothesised that subjects with newly diagnosed TIA would have evidence of an associated mild cognitive impairment; this being a manifestation of the same pathological process underlying the pathogenesis of the vascular event being initiated and accelerated by VRFs. The aims of the current study were, therefore, (i) to examine whether patients with first ever TIA and no history of stroke have evidence of cognitive impairment and, if so, whether the extent of the impairment was greater than expected compared with an age-, sex-matched control populations without VRFs and (ii) to determine which VRFs are associated with cognitive impairment. We conducted a case–controlled study between August and November 2008 in a University Hospital in UK (catchment population ∼750,000). Cases were defined as those patients with first ever TIA aged ≥45 years, assessed in a Neurovascular clinic, who satisfied the study inclusion criteria, i.e. a focal neurological deficit of <24 h duration of presumed vascular origin [13]. All diagnoses were confirmed by a highly experienced stroke physician unaware of the cognitive test results. Age (±1 year) and sex-matched controls with no known pre-defined VRFs or evidence of vascular disease were recruited from the dermatology and urology clinics at the same hospital. Neither TIA patients nor controls had known pre-existing cognitive impairment and/or depression. VRFs were defined as hypertension, hypercholesterolaemia, diabetes mellitus, atrial fibrillation, current smoker, previous history of myocardial infarction (MI) and a positive family history of cerebrovascular disease (TIA/stroke) based on the criteria set in the British Hypertension Society guidelines [14]. Ethical approval was obtained from the local ethics committee. For detailed recruitment methods please see Supplementary data are available Age and Ageing online, Appendix S1. To be able to detect at least a 20% difference in overall MoCA scores with 80% power at the 5% significance level, we recruited 68 cases and 68 controls. The MoCA was used as it is a more sensitive tool than MMSE in detecting mild cognitive decline [15, 16] and it also helps identify those who at risk of developing dementia [17]. Data were analysed using SPSS for Windows Version 16.0 (SPSS Co., Chicago, IL, USA) and presented descriptively. A group comparison was made between two groups for total as well as individual domain scores using unpaired Student's t-test. Presence or absence of association between cognitive impairment (MoCA <26) [8] for individual VRFs and a number of VRFs were examined using Chi-squared test. A total of 136 patients were recruited (146 approached, 10 (5 in each group) declined), 60% male (mean = 74.0 years). Table 1 shows the sample characteristics. As a result of selection criteria, hypertension and diabetes were more prevalent in TIA patients though both groups had similar levels of education and all were Caucasians. Thirty-nine (57%) of the TIA patients had cognitive impairment (MoCA <26) compared with none of the controls. The mean MoCA scores were significantly different between cases and controls (P < 0.0001). Controls had significantly higher scores than TIA patients in visuo-constructional/spatial skills, attention and concentration, calculation, language, memory, conceptual thinking and orientation domains (P < 0.0001 for all) but no difference was observed in the naming domain (P = 0.30) (see Table 2). Sub-group comparison of TIA patients who had impaired cognition (MoCA <26; n = 39) and those who did not (n = 29) showed current smokers, and who have had a previous MI were more likely to be cognitively impaired. Patients with three or more risk factors had a worse MoCA score than those with no VRFs (P = 0.03). Sample characteristics of TIA and non-TIA patients Sample characteristics of TIA and non-TIA patients Montreal cognitive assessment (MoCA) total and individual domain scores in TIA and non-TIA patients Montreal cognitive assessment (MoCA) total and individual domain scores in TIA and non-TIA patients We found that TIA patients with VRFs had worse cognitive function than age-/sex-matched non-TIA patients without VRFs and that there appears to be an association between the likelihood of cognitive impairment with an increasing number of VRFs. Previous studies have focused mainly on the relationship between either stroke or VRFs [3, 18] and decline in cognitive functioning [2, 3, 19–21] but not in TIA. To date it is unclear how much of the observed association between cerebrovascular disease and cognitive dysfunction is mediated by VRFs or whether there is a direct causal relationship. Desmond et al. [22] showed that diabetes mellitus was a significant independent correlate of abstract reasoning deficits and visuo-spatial dysfunction while hypercholesterolaemia was a significant independent correlate of memory dysfunction. Birns et al. [18] found that modest reductions in BP (<5/3 mmHg) were associated with improvements in MMSE score and performance on immediate and delayed logical memory tasks. Another review [23], however, found no convincing evidence that lowering BP prevents the development of dementia or cognitive impairment in hypertensive patients without apparent prior cerebrovascular disease, but noted some evidence that midlife hypertension but not late life hypertension was related to cognitive decline. The difficulty in disentangling the impact of cerebrovascular disease and VRF in the association with cognitive dysfunction could be that they are all interrelated, and that the mechanisms underlying the development of the TIA and of the cognitive impairment were manifestations of the single pathological process and both serving as markers of severity of cerebrovascular damage and/or further risk. It is not clear what the effect of VRF treatment on cognitive function is, i.e. whether it is the VRF or its treatment that is associated with cognitive decline. The evidence to date suggests that cognitive function is more likely to be determined by the severity of VRF [24]. However, the relative contribution of the severity and duration of the existing VRFs and the reversibility of pathological process responsible for cognitive impairment associated with VRFs is virtually unknown. In this study only a previous history of MI and current smoking habit appeared to predict cognitive impairment. It may be that hypertension and diabetes act as intermediate risk factors and potentiated by current smoking though the study may have been underpowered to detect weaker relations. In the current report, cognitive assessments were conducted by a single assessor and, therefore, were standardised. We used the MoCA which has been shown to be the most specific and sensitive than other scoring systems such as MMSE in detecting early cognitive decline [25, 26]. Our study has limitations. This is a case–controlled study; the results being based on cross-sectional data, and, therefore, we were not able to infer any causal relationship. In both groups, no neuro-imaging studies were performed, therefore we are not able to rule out the possibility of undiagnosed stroke, as well as not able to take into account the underlying burden of small vessel disease and to look for alternative explanations of cognitive impairment, for example selected hippocampal atrophy. However, the local practice pattern is that any patients' transient neurological deficit longer than an hour routinely receives CT or MRI imaging and hence this possibility is negligible. We were unable to adjust for duration or severity of risk factors, effects of current treatment on cognition and the level of risk factor control. Some previously unrecognised risk factors that were identified at the time of TIA clinic attendance were also not considered. Our sample size was only adequate to detect overall group differences between cases and controls. The study was under-powered to compare VRFs between subgroups. Smoking was the most prevalent VRF (74%) in TIA patients but other risk factors were less prevalent, e.g. AF. Therefore the question remains—does TIA confer any additional risk of cognitive impairment compared with that of the established VRFs? Other limiting factors include the non-blinding of cognitive assessment. The MoCA though sensitive screening tool lacks the value of a more detailed psychometric assessment particularly for frontal-executive functioning that may be particularly pertinent in early vascular-related cognitive impairment. We also did not assess possible depression. Our findings may not be directly applicable to other populations being based wholly on a UK Caucasian population. This study has shown that TIA patients had significant cognitive impairment compared with controls without VRF at the time of diagnosis. At the moment cognitive status is not routinely assessed in TIA patients. It is desirable to detect such early cognitive decline as it carries a poor prognosis [11]. Further studies with larger sample sizes in different patient populations are required prior to recommending routine cognitive assessment in TIA work-up. Whether optimal management of VRF would prevent decline or improve cognitive function in TIA patients with VRFs requires further evaluation. TIA patients with VRFs had worse cognitive function compared to age-/sex-matched controls without VRFs. The cognitive impairment in TIA appears to be associated with the number of VRFs. Whether optimal management of VRF would prevent decline or improve cognitive function in TIA patients with VRFs requires further investigations. None declared. Dr Guyomard was funded by the Academic Foundation Programme, Eastern Deanery. The study is supported by the Department of Medicine for the Elderly, Norfolk and Norwich University Foundation Hospital NHS Trust, Norwich, Norfolk, UK. Supplementary data mentioned in the text is available to subscribers in Age and Ageing online. We would like to thank the participants of the study. We gratefully acknowledge the Clinical Directors of Departments of Medicine for the Elderly, Dermatology and Urology, and staff from the TIA, Dermatology and Urology Clinics for their support and contribution. We would like to thank Dr Stanton, Consultant Neurologist, who suggested the use of the MoCA (Montreal Cognitive Assessment) assessment tool in our research and Dr Ziad Nasreddine, who authorised us to use it. We also thank Mrs Linda Davidson and Niki Day, Department of Medicine for the Elderly Secretaries, Norfolk and Norwich University Hospital for their secretarial support including proof reading.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.